Ting Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ting Wang has a total of 147 co-authors. Here are the top co-authors by number of shared publications:
- Mounir Hamdi
Hamad Bin Khalifa University
21 shared publications - Yuanming Shi
ShanghaiTech University
17 shared publications Yu Xia
14 shared publicationsZhiyang Su
The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
13 shared publications- Mingsong Chen
East China Normal University, Shanghai, China
11 shared publications - Haibin Cai
East China Normal University
8 shared publications
Latest Publications
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (12) , 22762-22779
Generative Diffusion-Based Black-Box Optimization for Robust Information Bottleneck Representation Learning
- 1 December 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (12) , 22762-22779
Generative Diffusion-Based Black-Box Optimization for Robust Information Bottleneck Representation Learning
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (17) , 40522-40537
DMT-PPO: Weight-Adaptive Multiobjective Task Offloading With Dynamic Preference Learning in Heterogeneous Edge Computing
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (17) , 40522-40537
DMT-PPO: Weight-Adaptive Multiobjective Task Offloading With Dynamic Preference Learning in Heterogeneous Edge Computing
- 28 July 2026
- Published by Elsevier in Journal of Energy Storage
- Vol. 179
A large language model approach with physics-guided data augmentation for early warning of battery thermal runaway
- 28 July 2026
- Published by Elsevier in Journal of Energy Storage
- Vol. 179
A large language model approach with physics-guided data augmentation for early warning of battery thermal runaway
- 18 July 2026
- Published by Springer Nature
- p. 404-415
Elite-Smoothing Particle Swarm Optimization for Efficient UAV Path Planning
- 18 July 2026
- Published by Springer Nature
- p. 404-415
Elite-Smoothing Particle Swarm Optimization for Efficient UAV Path Planning
- 14 July 2026
- Published by Springer Nature
- p. 458-469
Hybrid Quantum-MambaVision: A Quantum-Enhanced State Space Model for Calibrated Mixed-Type Wafer Defect Detection
- 14 July 2026
- Published by Springer Nature
- p. 458-469
Hybrid Quantum-MambaVision: A Quantum-Enhanced State Space Model for Calibrated Mixed-Type Wafer Defect Detection
- Open Access
- 11 April 2026
- Published by Elsevier in SSRN Electronic Journal
Physics-Informed LLM-based Data Augmentation and Transformer-based Early Warning of Battery Thermal Runaway
Battery thermal runaway (TR) represents a critical safety threat to electrified transportation systems. However, developing reliable early-warning models using onboard telemetry is highly challenging due to the extreme scarcity of validated TR precursor samples and the resulting severe class imbalance. To address this issue, this study proposes a physics-informed TR early-warning framework that integrates large language model (LLM)-based time-series data augmentation with a Transformer-based predictor.Specifically, the framework first generates candidate multichannel precursor windows through structured prompting. A multi-stage curation pipeline is then employed, incorporating hard physical constraints, distributional compatibility evaluation, heuristic coupling rules, and physics-informed residual screening derived from simplified thermo-electrochemical relationships. The curated samples are used to augment the minority class, while the same residual functional is further embedded as an auxiliary physics-guided regularizer during model training, forming a closed-loop consistency between data construction and model optimization.Experiments conducted on real-world fleet telemetry from 64 electric vehicles, including four vehicles with confirmed TR incidents, demonstrate that the proposed framework significantly improves precursor detection performance compared with representative baseline methods. In particular, the physics-constrained augmentation and regularization consistently improve minority-class recall and F1-score compared with models trained on real data alone or on LLM-generated data without physical screening. These results highlight the practical effectiveness of enforcing physical plausibility in rare-event trajectory generation and learning for battery safety monitoring.
- 11 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessPhysics-Informed LLM-based Data Augmentation and Transformer-based Early Warning of Battery Thermal Runaway
Battery thermal runaway (TR) represents a critical safety threat to electrified transportation systems. However, developing reliable early-warning models using onboard telemetry is highly challenging due to the extreme scarcity of validated TR precursor samples and the resulting severe class imbalance. To address this issue, this study proposes a physics-informed TR early-warning framework that integrates large language model (LLM)-based time-series data augmentation with a Transformer-based predictor.Specifically, the framework first generates candidate multichannel precursor windows through structured prompting. A multi-stage curation pipeline is then employed, incorporating hard physical constraints, distributional compatibility evaluation, heuristic coupling rules, and physics-informed residual screening derived from simplified thermo-electrochemical relationships. The curated samples are used to augment the minority class, while the same residual functional is further embedded as an auxiliary physics-guided regularizer during model training, forming a closed-loop consistency between data construction and model optimization.Experiments conducted on real-world fleet telemetry from 64 electric vehicles, including four vehicles with confirmed TR incidents, demonstrate that the proposed framework significantly improves precursor detection performance compared with representative baseline methods. In particular, the physics-constrained augmentation and regularization consistently improve minority-class recall and F1-score compared with models trained on real data alone or on LLM-generated data without physical screening. These results highlight the practical effectiveness of enforcing physical plausibility in rare-event trajectory generation and learning for battery safety monitoring.
- Open Access
- 10 April 2026
- Published by Elsevier in Engineering
- Vol. 63, 278-286
Generative and Large AI Models for 6G Wireless Networks: The Optimization Perspective
- 10 April 2026
- Published by Elsevier in Engineering
- Vol. 63, 278-286
Open AccessGenerative and Large AI Models for 6G Wireless Networks: The Optimization Perspective
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (6) , 9284-9299
Federated Linear Bandit Learning via UAV Aided Over-the-Air Computation
This paper investigates federated contextual linear bandit learning in a wireless network with a central server and multiple devices. To reduce communication latency, devices interact with the server via over-the-air computation (AirComp) over noisy, fading channels, where signal distortion can occur due to channel imperfections. Departing from traditional AirComp designs for static networks, we propose a novel federated bandit learning framework that leverages uncrewed aerial vehicles (UAVs) as mobile servers to aggregate data from distributed IoT devices. To optimize this system, we employ a block coordinate descent method combined with the alternating direction method of multipliers (BCD-ADMM), jointly optimizing the UAV trajectory, receive normalization factor, and transmission power to minimize the time-averaged mean square error (MSE) of AirComp. Our approach addresses the challenge of decentralized data across multiple devices, enabling secure and efficient collaboration without direct data sharing. Theoretical analysis establishes an upper bound on the algorithm’s regret, affirming the framework’s scalability and robustness against noise. Simulation results support these findings, highlighting notable performance improvements in federated bandit learning with UAV-assisted AirComp.
- 12 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (6) , 9284-9299
Federated Linear Bandit Learning via UAV Aided Over-the-Air Computation
This paper investigates federated contextual linear bandit learning in a wireless network with a central server and multiple devices. To reduce communication latency, devices interact with the server via over-the-air computation (AirComp) over noisy, fading channels, where signal distortion can occur due to channel imperfections. Departing from traditional AirComp designs for static networks, we propose a novel federated bandit learning framework that leverages uncrewed aerial vehicles (UAVs) as mobile servers to aggregate data from distributed IoT devices. To optimize this system, we employ a block coordinate descent method combined with the alternating direction method of multipliers (BCD-ADMM), jointly optimizing the UAV trajectory, receive normalization factor, and transmission power to minimize the time-averaged mean square error (MSE) of AirComp. Our approach addresses the challenge of decentralized data across multiple devices, enabling secure and efficient collaboration without direct data sharing. Theoretical analysis establishes an upper bound on the algorithm’s regret, affirming the framework’s scalability and robustness against noise. Simulation results support these findings, highlighting notable performance improvements in federated bandit learning with UAV-assisted AirComp.
- 4 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (5) , 7196-7209
Throughput-Optimized Service Routing for Microservice Flows in LEO Satellite Networks
Satellite-based microservice systems have emerged as a promising architecture for enabling scalable and flexible service deployment in Low Earth Orbit (LEO) satellite networks, which are increasingly recognized as a key solution to meet the rising demand for global communication and computation, especially in remote and underserved regions. However, routing microservices efficiently in such systems presents major challenges due to the dynamic topology, intermittent connectivity, and unstable link conditions inherent to satellite constellations. These issues become even more severe under high traffic loads. To address these challenges, we propose Service Pressure, a novel routing algorithm specifically designed for satellite-based microservice systems. Service Pressure comprises two key components: first, the construction of an augmented subgraph to model the complex execution and data transmission dependencies of microservices; second, a distributed service routing algorithm that utilizes queue backlogs. This combination enables the algorithm to effectively handle high throughput and adapt to the dynamic network conditions of satellite constellations. By optimizing resource utilization, minimizing latency, and balancing load across satellite nodes, Service Pressure ensures efficient and stable service orchestration even under fluctuating traffic conditions. Extensive simulations demonstrate that our approach significantly outperforms existing routing methods, particularly in terms of throughput, latency, and stability. Service Pressure offers a significant advancement in satellite microservice routing, making it ideal for next-generation space-ground integrated networks.
- 4 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 25 (5) , 7196-7209
Throughput-Optimized Service Routing for Microservice Flows in LEO Satellite Networks
Satellite-based microservice systems have emerged as a promising architecture for enabling scalable and flexible service deployment in Low Earth Orbit (LEO) satellite networks, which are increasingly recognized as a key solution to meet the rising demand for global communication and computation, especially in remote and underserved regions. However, routing microservices efficiently in such systems presents major challenges due to the dynamic topology, intermittent connectivity, and unstable link conditions inherent to satellite constellations. These issues become even more severe under high traffic loads. To address these challenges, we propose Service Pressure, a novel routing algorithm specifically designed for satellite-based microservice systems. Service Pressure comprises two key components: first, the construction of an augmented subgraph to model the complex execution and data transmission dependencies of microservices; second, a distributed service routing algorithm that utilizes queue backlogs. This combination enables the algorithm to effectively handle high throughput and adapt to the dynamic network conditions of satellite constellations. By optimizing resource utilization, minimizing latency, and balancing load across satellite nodes, Service Pressure ensures efficient and stable service orchestration even under fluctuating traffic conditions. Extensive simulations demonstrate that our approach significantly outperforms existing routing methods, particularly in terms of throughput, latency, and stability. Service Pressure offers a significant advancement in satellite microservice routing, making it ideal for next-generation space-ground integrated networks.